This document is downloaded at: 2018-06-05T09:32:07Z

Title A novel and specific fluorescence reaction for uracil.

Author(s) Shibata, Takayuki; Kawasaki, Shin-ya; Fujita, Jun-ya; Kabashima,Tsutomu; Kai, Masaaki

Citation Analytica chimica acta, 674(2), pp.234-238; 2010

Issue Date 2010-08-03

URL http://hdl.handle.net/10069/23653

Right Copyright © 2010 Elsevier B.V. All rights reserved.

NAOSITE: Nagasaki University's Academic Output SITE

http://naosite.lb.nagasaki-u.ac.jp

1

A novel and specific fluorescence reaction for uracil

Takayuki Shibataa, Shin-ya Kawasakia, Jun-ya Fujitaa, Tsutomu Kabashimaa, Masaaki Kaia,b,*

aFaculty of Pharmaceutical Sciences, Graduate School of Biomedical Sciences, Nagasaki

University, 1-14 Bunkyo-Machi, Nagasaki 852-8521, Japan bGlobal Center of Excellence Program, Nagasaki University

*Corresponding author. Tel. and fax: +81-95-819-2438; E-mail: ms-kai@nagasaki-u.ac.jp.

ABSTRACT

Facile and specific methods to quantify a nucleobase in biological samples are of great

importance for diagnosing disorders in nucleic acid metabolism. In the present study, a novel

fluorogenic reaction specific for uracil has been developed. The reaction was carried out in an

alkaline medium containing benzamidoxime and K3[Fe(CN)6] with heating for 2.0 min.

Under the optimum reaction conditions, strong fluorescence was produced only from uracil,

not from other many biogenic compounds tested such as cytosine, thymine, adenine, guanine,

nucleobases, nucleosides, nucleotides, amino acids, saccharides, creatine, creatinine and urea.

The sensitivity of this method was compared with a known fluorogenic reaction using

phenacylbromide which does not react with uracil but reacts with cytosine, adenine and their

analogues. The proposed uracil-specific reaction showed approximately 400 fold higher

sensitivity than the phenacylbromide reaction. The lower detection limit of uracil by the

present method was 100 pmol mL-1, and a good linearity of the calibration curve was obtained

up to 100 nmol mL-1 uracil. Due to its high sensitivity and specificity, the quantitative

determination of uracil was possible by the proposed fluorimetric method.

Keywords: Fluororescence reaction; Benzamidoxime; Uracil specificity; High sensitivity;

Pyrimidine metabolism.

2

1. Introduction

Uracil is metabolised to -alanine via a three-step catabolic pathway which includes the

rate-determining step from uracil to 5, 6-dihydrouracil by dihydropyrimidine dehydrogenase

(DPD) in a mammalian body, as shown in Fig. 1 [1-4]. On the other hand, fluorinated uracil

derivatives such as 5-fluorouracil (5-FU) and its pro-drug, 5-fluoro-1-(tetrahydro-2-furanyl)-

2,4-(1H,3H)-pyrimidine-dione (tegafur), can be potent anticancer drugs because of the

inhibition of DNA replication [5]. Since 5-FU also is catabolised by DPD [6-8], a loss of DPD

activity causes an accumulation of 5-FU in blood after administration, and leads to severe side

effects including death (Fig. 1, right side) [9-14]. Moreover, inherited disorders in the

degradation pathway of pyrimidine are the major criteria of the inborn errors of pyrimidine

metabolism, and are known to show a variety of symptoms such as convulsion, mental

retardation and microcephaly [15]. Quantification of uracil in blood and urine is of great

importance to identify the level of pyrimidine metabolism, especially for the diagnosis of

DPD deficiency.

Fig. 1. Catabolic pathway of uracil and 5-FU.

Methods for the quantification of uracil in biological samples have utilized

high-performance liquid chromatography (HPLC) and mass spectrometry (MS) e.g. tandem

HPLC method with column switching [16], HPLC/tandem MS system [17], gas

3

chromatography/MS system [18] and radiometric HPLC [19-22]. Although these methods are

superior for the determination of a complex sample, the instrumentations need high cost for

the analysis. Recently, Mattison et al. demonstrated a non-chromatographic method for the

analysis of the level of uracil degradation, in which 2-[13C]-labelled uracil was administrated

and the catabolised 13CO2 in exhaled breath sample was measured by infrared spectrometry

[23]. This method is convenient for evaluating the metabolic function of uracil. However, the

stage at which the error occurs in pyrimidine catabolism cannot be identified since three

enzymes are involved in the digestion of uracil to CO2.

Therefore, a facile and high-throughput technique to quantify uracil has been highly

demanded. Fluorescence (FL) reaction is one of the widely-used methodologies for sensitive,

facile and simultaneous detection of multiple samples in a multi-well plate. In this article,

we describe a novel FL-deriving reaction specific for uracil for the first time. This reaction

was conducted by heating uracil with benzamidoxime (BAO) and K3[Fe(CN)6] in a KOH

solution. The proposed method could sensitively measure uracil with a lower detection limit

of 50 fmol per an injection volume by HPLC and 100 pmol mL-1 in the reaction mixture by a

spectrofluorimeter. To highlight the specificity of the current reaction, the mechanism of the

reaction is also discussed.

2. Experimental

2.1. Chemicals

Benzamidine hydrochloride and K3[Fe(CN)6] were purchased from Nacalai Tesque (Kyoto,

Japan). The other oxidising agents were from Wako (Osaka, Japan). BAO and

benzohydroxamic acid were obtained from TCI (Tokyo, Japan). The other BAO-related

chemicals were from Sigma-Aldrich (MO, USA). All chemicals were of analytical or

guaranteed reagent grade and were used without further purification. Water was purified on

a Milli-Q system WR 600 A from Millipore (Molsheim, France).

2.2. Procedure recommended for the FL reaction of uracil

Uracil, BAO, K3[Fe(CN)6] and KOH were weighed in different flasks and water was

added to them to prepare for four stock solutions. To a 1.5-mL tube was added 250 L of each

reactant in the order of 4.0ｘ10-4 - 4.0 mol mL-1 uracil, 4.0 mmol L-1 BAO, 8.0 mmol L-1

K3[Fe(CN)]6 and 4.0 mol L-1 KOH. The mixture was then heated at 90 ˚C for 2.0 min,

followed by cooling in an ice bath for 2.0 min to stop the reaction.

2.3. FL detection

The relative FL intensity of the final reaction mixture in a quartz cell (10 × 4 mm) was

measured with an FP-6300 spectrofluorimeter (Jasco, Tokyo, Japan) at excitation and

4

emission wavelengths of 320 nm and 410 nm, respectively with both the band widths of 5 nm.

The identification of the fluorescent derivative of the nucleobase was carried out on a

reversed-phase HPLC system, consisting of a PU-2089 quaternary gradient pump, an FP-920

FL detector, and an UV-970 detector (Jasco, Tokyo, Japan). The separation of the fluorescent

derivative was performed on a TSKgel ODS-80Ts reversed-phase column (250 × 4.6-mm ID)

(TOSOH, Tokyo, Japan) with an isocratic elution employing 30 % (v/v) methanol in water

containing 10 mmol L-1 triethylammonium acetate (pH 7.0) at a constant flow rate of 1.0 mL

min-1.

2.4. Phenacylbromide reaction

The reaction was conducted according to the reported procedure with a slight modification

[24]. To a 200-L aqueous solution of cytosine in a 1.5-mL tube was added 600 L of acetic

acid-acetonitrile-water solution (1:80:19, v/v) followed by adding 200 L of 12.5 mmol L-1

phenacylbromide in acetonitrile. The reaction mixture was heated at 80 ˚C for 45 min. The FL

intensity was then measured in the same manner as described above, except for an excitation

and emission wavelengths of 320 nm and 370 nm, respectively.

2.5. Large scale synthesis of FL product

To a 25-mL portion of 1.0 mol mL-1 uracil (2.8 mg, 25 mol) was added 25 mL each of

4.0 mmol L-1 4-trifluoromethylbenzamidoxime (20.4 mg, 0.1 mmol), 8.0 mmol L-1

K3[Fe(CN)6] (65.9 mg, 0.2 mmol) and 2.0 mol L-1 KOH. All the chemicals were dissolved in

H2O. The mixture was heated at 100˚C for 10 min, at which point the reaction was cooled in

an ice bath. Four identical reactions were conducted simultaneously, and the resulting reaction

mixtures (4 x 100 mL) were combined. Potassium chloride was added to the mixture until it

reached saturation, and the remaining precipitate of KCl was filtered off. An aqueous layer

was extracted with ethylacetate (4 times with 100 mL), and the combined organic layer was

dried over Na2SO4, filtered and evaporated to dryness to give 45 mg of the crude product. The

above procedure was repeated for 13 times and the resulting crude products were combined

(826 mg). Purification by silica gel column chromatography (ethylacetate : methanol = 9:1 to

8:2 to 7:3) gave the analytically pure product as a pale blue powder (25 mg, 6 %).

3. Results and discussion

3.1. Reaction of uracil and cytosine with BAO.

In our preliminary studies to develop a new fluorogenic reaction specific for uracil, we

found that uracil emitted blue FL when it was heated with BAO and K3[Fe(CN)6] in an

alkaline aqueous solution. When uracil was replaced with cytosine, weak blue FL was

observed. The other three bases (adenine, guanine and thymine), uridine and 2’-deoxyuridine

were all fluorescent-inactive. BAO is a non-fluorescent compound and has not ever been

5

utilized for FL reactions. Furthermore, there was no FL reaction specific for uracil. Thus, the

discovery of this unique FL reaction encouraged us to further investigate characteristics of the

reaction.

In order to increase the specificity for uracil, this reaction conditions were investigated

using 250 nmol mL-1 each of uracil and cytosine (Fig. 2). FL intensities were produced from

uracil and cytosine, and both increased with the concentration of K3[Fe(CN)6] up to 2 mmol

L-1 (Fig. 2a). However, the FL was dramatically decreased when the concentration of the

oxidant was further increased. Influence of the BAO concentration (Fig. 2b) indicated

different pattern from the results for the K3[Fe(CN)6] concentration. No FL was initially

generated up to 0.75 mmol L-1 of the BAO concentration. After further increase of the BAO

concentration, then dramatic increases of the FL intensity were observed.

Although similar patterns for the uracil and cytosine conditions were obtained by the

reaction temperature (Fig. 2d) as well as the concentrations of K3[Fe(CN)6] and BAO, KOH

concentration (Fig. 2c) and reaction time (Fig. 2e) showed different patterns between the

uracil and cytosine reactions. Non FL was produced when the reaction was conducted without

KOH, however the FL was generated by increasing the concentration of KOH. In case of

uracil, a maximum FL intensity was obtained when the KOH concentration was 0.5 mol L-1 or

greater, whilst 0.025 mol L-1 KOH was found to be the optimum for cytosine. In difference

of the reaction time, the FL of cytosine was not produced until 2 min (Fig. 2e). In case of

uracil, however its FL intensity gradually increased with reaction time and reached the

maximum at 5 min. These results might be attributed to the difference in the reaction

mechanism between uracil and cytosine.

In order to investigate a process of the FL production, each reaction mixture for uracil and

cytosine was analysed by HPLC with FL detection (Fig. 2 f and g). The HPLC profile for the

reaction mixture of uracil showed a single FL peak, and a lower detection limit of uracil was

approximately 50 fmol per injection (S/N=3). On the other hand, two small FL peaks

appeared in case of cytosine, and one of two peaks was at the same retention time as that of

the uracil product. The results suggested that the initiation of the reaction with cytosine relies

on the decomposition of cytosine, essentially deamination at the 4-position of cytosine to

uracil.

6

Fig. 2. Reaction conditions for uracil (-●-) and cytosine (-▲-). (a) K3[Fe(CN)6] concentration,

(b) BAO concentration, (c) KOH concentration, (d) reaction temperature and (e) reaction time.

The reaction mixtures with (f) uracil and (g) cytosine were individually separated by HPLC

with FL detection.

Table 1 Effect of the order of reagent addition on FL intensity.

The order of reagent additiona RFI d The order of reagent additiona RFI d

U→B→F→K 100 F→K→U→B 76

U→B→K→F 91 F→K→B→U 24

U→F→B→K 98 K→U→B→F 90

U→F→K→B 95 K→U→F→B 91

U→K→B→F 92 K→B→U→F 92

U→K→F→B 92 K→B→F→U 22

B→U→F→K 54 K→F→U→B 80

B→U→K→F 64 K→F→B→U 28

B→F→U→K 104 (U+B)→(F+K)b 89

B→F→K→U 14 (U+F)→(B+K)b 92

B→K→U→F 109 (U+K)→(B+F)b 92

B→K→F→U 56 (U+B+F)→Kc 86

F→U→B→K 85 (U+B+K)→Fc 92

F→U→K→B 83 (U+F+K)→Bc 85

7

F→B→U→K 93 (B+F+K)→Uc 12

F→B→K→U 26 a Each stock solution of uracil (U), BAO (B), K3[Fe(CN)6] (F) and KOH (K) was added in the

order from the left to the right. b Two components of the four reagents were mixed and left standing for 10 min prior to the

reaction. c Three components of the four reagents were mixed and left standing for 10 min prior to the

reaction. d The RFI obtained by the reaction was taken as 100, in which the reactants were added in the

order of uracil, BAO, K3[Fe(CN)6] and KOH.

The order in which the reactants were mixed was also seen to affect the results of the

reaction (Table 1). After testing 24 combinations, we found that the disordered addition of the

reagents weakened the FL intensity, especially when uracil was added last. If any two

reactants were mixed and subsequently added to the other two components, there was no

diminishing of the FL. However, inhibition of the FL production was observed when BAO,

K3[Fe(CN)6] and KOH were mixed and then incubated prior to the addition of uracil. These

results indicated that two major reaction pathways, namely the production of the FL

compound and the decomposition of reactants, exist in the present reaction. In order to

increase the reaction efficiency, it was better that each reactant was mixed in the order of

nucleobase, BAO, K3[Fe(CN)6] and KOH.

3.2. Effect of reactants

BAO (Fig. 3a) was replaced by several BAO-related compounds containing amidoxime

functional group. BAO was found to be the most effective fluorogenic reagent, and

substitution at any position of phenyl ring suppressed the production of FL (Fig. 3).

4-Trifluoromethylbenzamidoxime (4-trifluoromethyl-BAO) (Fig. 3d) and

4-trifluoromethoxy-BAO (Fig. 3e) generated weaker FL, however, neither

3-trifluoromethyl-BAO (Fig. 3b) nor 3,5-bis(trifluoromethyl)-BAO (Fig. 3c) produced

measurable FL. Since amidoxime group is the most reactive moiety in BAO-related

compounds, substitution on the phenyl ring would affect the reactivity of amidoxime. The

result that formamidoxime (Fig. 3f) did not produce FL suggested that the phenyl ring of

BAO is involved in the fluorophore of the product. Non FL was observed when benzamidine

(Fig. 3g) was used, whereas benzohydroxamic acid (Fig. 3h) gave a very weak FL. The results

demonstrate that the amidoxime group is important to produce the FL substance. A

tautomerism of benzohydroxamic acid results in the presence of oxime and hydroxyl groups,

which is a similar structure to BAO. This unfavorable tautomerism and weak nucleophilicity

of the hydroxyl group would decrease the rate of the conjugation between benzohydroxamic

8

acid and uracil. However, even a very small amount of the coupled intermediate might cause a

further reaction between the uracil moiety and the amidoxime group, which produced the

fluorophore in the same manner as the BAO reaction. Hence the non-fluorogenicity of

benzamidine should be attributed to the lack of the reactivity at second stage of the

conjugation reaction due to the absence of the amidoxime group. These results indicate that

the fluorogenic reagents must have 1) an oxime (or hydroxylamino) group, 2) a nucleophilic

functional group next to oxime, and 3) aromaticity. The substitutions on the phenyl ring also

affect the fluorogenisity and might break the balance of the reactivity of 1) and 2).

Fig. 3. Reactions of uracil with several BAO-related compounds. (a) BAO, (b)

3-trifluoromethyl-BAO, (c) 3,5-bis(trifluoromethyl)-BAO, (d) 4-trifluoromethyl-BAO, (e)

4-trifluoromethoxy-BAO, (f) formamidoxime, (g) benzamidine and (h) benzohydroxamic

acid.

3.3. Reaction specificity

The reaction should give a sufficient FL intensity for uracil without any FL intensity for

cytosine. According to Fig. 2, the FL production of cytosine was diminished by either

increasing the KOH concentration or shortening the reaction time. Such conditions were

obtained when the final concentrations of 1.0 mmol L-1 BAO, 2.0 mmol L-1 K3[Fe(CN)]6 and

1.0 mol L-1 KOH were adopted and its mixture was heated at 90 ˚C for 2 min.

Specificity of this reaction was then checked using other nucleobases (cytosine, thymine,

adenine and guanine), nucleosides (2’-deoxyuridine, 2’-deoxycytidine, 2’-deoxythymidine,

9

2’-deoxyadenosine and 2’-deoxyguanosine), nucleotides (5’-monophosphate of uridine,

cytidine, thymidine, adenosine and guanosine) and several nucleobase analogues

(1-methyluracil, 5-fluorouracil, pseudouridine, 6-methyluracil and 5, 6-dihydrouracil). As

shown in Fig. 4, uracil only provided FL amongst all substances tested. On the basis of the

above results, it is suggested that one or more substitutions at 1-, 5- or 6-position of uracil

can disturb the conversion into the FL product, so that uracil-containing nucleosides and

nucleotides did not produce the FL compounds. In addition, this reaction did not provide any

FL products for 20 kinds of amino acid, sugars (glucose, fructose, lactose, ribose and

sucrose), creatine, creatinine and urea. Thus, potential of the present reaction is greatly

intense for the accurate quantification of uracil in biological samples without influences

from other nucleobases and nucleobase-related components.

Fig. 4. Reaction specificity of the present method. A final concentration of 250 nmol mL-1 of

each substance was used except for amino acids (a mixture of 20 amino acids consisted of 10

mM of each amino acid).

3.4. Comparative study

Phenacylbromide is one of few fluorogenic reagents selective for nucleobases [24]. Since

no fluorogenic reaction specific for uracil has been reported, the sensitivity of the present

BAO reaction was compared with the phenacylbromide reaction using cytosine. The result

showed that the BAO reaction with uracil was approximately 400 times more sensitive than

the phenacylbromide reaction with cytosine. In case of the phenacylbromide reaction, a linear

correlation (Y = 1.44×107 X + 3.28, r2=0.994; Y is the FL intensity and X is the concentration,

mol mL-1 of cytosine) between cytosine concentration and FL intensity was obtained up to 2.5

mol mL-1. In contrast, even 1.0 nmol mL-1 uracil could be unambiguously detected by the

BAO reaction (Fig. 5). The lower detection limit of uracil by the present method was found to

10

be approximately 100 pmol mL-1 (S/N=3), and a good linearity (Y = 6.27×109 X + 0.31,

r2=0.999; Y is the FL intensity and X is the concentration, mol mL-1 of uracil). was obtained

up to 100 nmol mL-1 uracil It is known that phenacylbromide reacts not only with cytosine but

also with adenine and nucleos(t)ides containing cytosine and adenine under heating for 45

min [24]. Therefore, the present FL reaction is rapid and sensitive, possessing an excellent

specificity for uracil.

Fig. 5. Comparative study between the reactions of BAO with uracil and of phenacylbromide

with cytosine. Nucleobases at 1.0 nmol mL-1 were treated with the BAO reaction (n=3, left

panel) and the phenacylbromide reaction (n=3, right panel), respectively. Blank refers to each

reaction in the absence of nucleobases.

3.5. Structural elucidation of FL product

Our next concern was to determine the chemical structure of the FL product as well as the

reaction mechanism. It was found that the FL product by the reaction of uracil with BAO was

water-soluble and hard to isolate via extraction from the aqueous solution. Therefore,

4-trifluoromethyl-BAO (Fig. 3d) was used for a large scale synthesis. A single FL product was

successfully obtained as a light-blue powder after extraction with ethyl acetate followed by

silica gel column chromatographic separation. 1H-NMR data showed that the product lost two

protons at the 5- and 6-positions of uracil, indicating that 4-trifluoromethyl-BAO covalently

attached to these positions while maintaining the double bond (data not shown). It is known

that the addition of carbonyl group at the C5 of uracil followed by oxidation of the resulting

hydroxyl group to a carbonyl group occurs under basic condition in the presence of an oxidant

[25]. The initiation of the reaction was thought to be the deprotonation at the N1 and the

nucleophilic attack to BAO at the C5. Therefore, substitution at either the N1 or the C5 would

inhibit the reaction, and thus this mechanism matches the result that the substitution at the N1

or the C5 did not produce the FL product. There are a couple of reports demonstrating the

oxidative cyclization of uracil derivatives containing hydroxiamino or nitroso group in the

presence of K3[Fe(CN)6] under basic condition [26, 27]. According to these reactions,

intramolecular cyclization most likely happens at the C6 with forming N-oxide, which suits

11

the result that the peak corresponding to -OH in the amidoxime group disappeared. The

reason why the reaction with 6-methyluracil did not proceed could be because the substitution

at the C6 of pyrimidine ring hinders the nitrogen atom from reaching close proximity due to

the planner structure of the intermediate. The predicted structure of the FL product between

uracil and 4-trifluoromethyl-BAO is shown in Fig. 6. An ESI-TOF mass spectrum showed

strong ion peaks at m/z = 297 ([M-oxigen+H]+), and relatively weak peak at m/z = 315

([M-oxigen+H2O+H]+) (data not shown). It is known that the compound containing N-oxide

shows a strong [M-oxigen+H]+ peak [27], which is the same phenomenon as the result

obtained in this study. In addition, the FL product by the reaction of uracil with BAO was

separated by HPLC (Fig. 2f), and the ESI-TOF mass spectrum was measured. Two main

peaks appeared at m/z = 229 ([M-oxigen+H]+) and 247 ([M-oxigen+H2O+H]+) (data not

shown) which were both 68 mass less than m/z = 297 and 315, respectively, obtained by the

reaction of uracil with 4-trifluoromethyl-BAO. The number of 68 mass corresponded to the

difference of the molecular mass between 4-trifluoromethyl-BAO and BAO. From the above

result, the reaction mechanism of uracil with BAO might be the same as that with

4-trifluoromethyl-BAO.

Fig. 6. Plausible reaction mechanism of the fluorogenic reaction between uracil and

4-trifluoromethyl-BAO and expected structure of the FL product.

4. Conclusions

In this work, we have developed the novel fluorogenic reaction specific for uracil. The

reaction was easily accomplished by heating the mixture of uracil and BAO in the presence of

K3[Fe(CN)6] under basic condition. Although the present system produced weak FL with

cytosine, the uracil-specific FL reaction could be achieved by controlling the reaction

conditions. The reaction was influenced by the structure of the fluorogenic reagent as well as

12

the reaction conditions. According to the analysis of the reaction mechanism, an important

factor for the pyrimidine substance to drive the reaction forward is to have no substitution at

either N1, C5 or C6 in the pyrimidine ring and carbonyl group at C4. Taking these things into

account, the proposed reaction shows severe structural requirements, indicative of the

superior specificity. The reaction gave more than 400 times stronger FL in shorter time as

compared with the conventional FL reaction of cytosine with phenacylbromide. These results

clearly corroborate that the BAO reaction affords the first example of a uracil-specific

fluorogenic reaction, and its highly sensitive and facile protocol should be utilized for further

analytical tools such as HPLC, capillary electrophoresis and mass spectrometry. The proposed

method will be applied to high-throughput determinations of uracil in urine or blood

specimens for a primary screening test of the inborn errors of pyrimidine metabolism such as

DPD deficiency.

Acknowledgements

This work was supported in part by the president’s discretionary fund of Nagasaki

University, Japan, and partly supported by the Global Center of Excellence Program in

Nagasaki University, Japan.

References

[1] B. Podschun, G. Wahler, K.D. Schnackerz, Eur. J. Biochem. 185 (1989) 219-224.

[2] T.W. Traut, S. Loechel, Biochemistry 23 (1984) 2533-2539.

[3] C. Wasternak, Pharm. Ther. 8 (1980) 629-651.

[4] M.M. Matthews, W. Liao, K.L. Kvalnes-Krick, T.W. Traut, Arch. Biochem. Biophys.

293 (1992) 254-263.

[5] A. Bronckaers, J. Balzarini, S. Liekens, Biochem. Pharmacol. 76 (2008) 188-197.

[6] G.D. Heggie, J.P. Sommadossi, D.S. Cross, W.J. Huster, R.B. Diasio, Cancer Res. 47

(1987) 2203-2206.

[7] D.H. Ho, L. Townsend, M.A. Luna, G..P. Bodey, Anticancer Res. 6 (1986) 781-784.

[8] T.M. Woodcock, D.S. Martin, L.A. Damin, N.E. Kemeny, C.W. Young, Cancer 45

(1980) 1135-1143.

[9] A.P. Lyss, R.C. Lilenbaum, B.E. Harris, R.B. Diasio, Cancer Invest. 11 (1993) 239-240.

[10] H. Okuda, T. Nishiyama, K. Ogura, S. Nagayama, K. Ikeda, S. Yamaguchi, Drug Metab.

Dispos. 25 (1997) 270-273.

[11] H. Okuda, K. Ogura, A. Kato, H. Takubo, T. Watabe, J. Pharmacol. Exp. Ther. 287

(1998) 791-799.

[12] T. Inada, T. Jotsuka, G.. Matsuda, H. Kawakubo, Y. Ogata, T. Kubota, Int. J. Clin. Oncol.

4 (1999) 54-56.

13

[13] A. Beck, M.C. Etienne, S. Chéradame, J.L. Fischel, P. Formento, N. Renée, G. Milano,

Eur. J. Cancer 30 (1994) 1517-1522.

[14] A. Lee, H. Ezzeldin, J. Fourie, R.B. Diasio, Clin. Adv. Hematol. Oncol. 2 (2004)

527-532.

[15] M. Duran, P. Rovers, P.K. de Bree, C.H. Schreuder, H. Beukenhorst, L. Dorland, R.J.

Berger, Inherit. Metab. Dis. 14 (1991) 367-370.

[16] S. Sumi, K. Kidouchi, S. Ohba, Y. Wada, J. Chromatogr. B 672 (1995) 233-239.

[17] H. van Lenthe, A.B.P. van Kuilenburg, T. Ito, A.H. Bootsma, A. van Cruchten, Y. Wada,

A.H. van Gennip, Clin. Chem. 46 (2000) 1916-1922.

[18] T. Kuhara, C. Ohdoi, M. Ohse, A.B.P. van Kuilenburg, A.H. van Gennip, S. Sumi, T. Ito,

Y. Wada, I. Matsumoto, J. Chromatogr. B 792 (2003) 107-115.

[19] B. Podschun, G. Wahler, K.D. Schnackerz, Eur. J. Biochem. 185 (1989) 219-224.

[20] R.A. Fleming, G. Milano, A. Thyss, M-C. Etienne, N. Renée, M. Schneider, F. Demani,

Cancer Res. 52 (1992) 2899 -2902.

[21] B.E. Harris, R. Song, S. Soong, R.B. Diasio, Cancer Res. 50 (1990) 197 -201.

[22] M.R. Johnson, J. Yan, L. Shao, N. Albin, R.B. Diasio, J. Chromatogr. B 696 (1997)

183-191.

[23] L.K. Mattison, H. Ezzeldin, M. Carpenter, A. Modak, M.R. Johnson, R.B. Diasio, Clin.

Cancer Res. 10 (2004) 2652 -2658.

[24] E.J. Eisenberg, K.C. Cundy, J. Chromatogr. B 679 (1996) 119-127.

[25] D.R. Hwang, P. Helquist, M.S. Shekhani, J. Org. Chem. 50 (1985) 1264-1271.

[26] A.A. Yavolovskii, E.I. Ivanov, Chem. Heterocyc. Compd. 40 (2004) 361-363.

[27] A.A. Yavolovskii, E.I. Ivanov, A.V. Mazepa, Y.E. Ivanov, Russ. J. Gen. Chem. 73 (2003)

1486-1488.

14

Figure captions

Fig. 1. Catabolic pathway of uracil and 5-FU.

Fig. 2. Reaction conditions for uracil (-●-) and cytosine (-▲-). (a) K3[Fe(CN)6] concentration,

(b) BAO concentration, (c) KOH concentration, (d) reaction temperature and (e) reaction time.

The reaction mixtures with (f) uracil and (g) cytosine were individually separated by HPLC

with FL detection.

Fig. 3. Reactions of uracil with several BAO-related compounds. (a) BAO, (b)

3-trifluoromethyl-BAO, (c) 3,5-bis(trifluoromethyl)-BAO, (d) 4-trifluoromethyl-BAO, (e)

4-trifluoromethoxy-BAO, (f) formamidoxime, (g) benzamidine and (h) benzohydroxamic

acid.

Fig. 4. Reaction specificity of the present method. A final concentration of 250 nmol mL-1 of

each substance was used except for amino acids (a mixture of 20 amino acids consisted of 10

mM of each amino acid).

Fig. 5. Comparative study between the reactions of BAO with uracil and of phenacylbromide

with cytosine. Nucleobases at 1.0 nmol mL-1 were treated with the BAO reaction (n=3, left

panel) and the phenacylbromide reaction (n=3, right panel), respectively. Blank refers to each

reaction in the absence of nucleobases.

Fig. 6. Plausible reaction mechanism of the fluorogenic reaction between uracil and

4-trifluoromethyl-BAO and expected structure of the FL product.